Conventional color digital cameras can only produce three-channel images so they are limited when high-quality color reproduction is required. Alternatively, spectral imaging increases the number of channels and can retrieve spectral reflectance for each scene pixel. The major goal of spectral imaging is high spectral accuracy, while it may also be beneficial to achieve high colorimetric accuracy for a specific viewing condition. A new spectral reconstruction method, called the matrix R method, was developed to achieve both goals simultaneously. An experiment was performed to test this method. The experimental results have been very promising; average color difference for all targets evaluated was about 1.3 CIEDE2000 and 2.0% RMS. These results suggest that this new method is a promising method for building digital image databases for museums, archives and libraries.
Department, Program, or Center
Chester F. Carlson Center for Imaging Science (COS)
Tenth Congress of the International Colour Association (2005) 469-472
RIT – Main Campus